| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 100, | |
| "global_step": 478, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 1.0416666666666666e-08, | |
| "logits/chosen": -0.3017902672290802, | |
| "logits/rejected": -0.4429064989089966, | |
| "logps/chosen": -319.77001953125, | |
| "logps/rejected": -256.397705078125, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 1.0416666666666667e-07, | |
| "logits/chosen": -0.6185635924339294, | |
| "logits/rejected": -0.6267194747924805, | |
| "logps/chosen": -266.4049072265625, | |
| "logps/rejected": -226.4547576904297, | |
| "loss": 0.6938, | |
| "rewards/accuracies": 0.5347222089767456, | |
| "rewards/chosen": -0.0314086489379406, | |
| "rewards/margins": 0.004176241811364889, | |
| "rewards/rejected": -0.035584889352321625, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 2.0833333333333333e-07, | |
| "logits/chosen": -0.5040268301963806, | |
| "logits/rejected": -0.5066824555397034, | |
| "logps/chosen": -294.06231689453125, | |
| "logps/rejected": -256.36572265625, | |
| "loss": 0.6919, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": 0.0255455132573843, | |
| "rewards/margins": 0.0022588702850043774, | |
| "rewards/rejected": 0.023286638781428337, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "logits/chosen": -0.47493982315063477, | |
| "logits/rejected": -0.4304034113883972, | |
| "logps/chosen": -294.8653869628906, | |
| "logps/rejected": -278.9495544433594, | |
| "loss": 0.6909, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": 0.011051720939576626, | |
| "rewards/margins": 0.003937236033380032, | |
| "rewards/rejected": 0.007114484906196594, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 4.1666666666666667e-07, | |
| "logits/chosen": -0.6350118517875671, | |
| "logits/rejected": -0.6409525871276855, | |
| "logps/chosen": -302.514404296875, | |
| "logps/rejected": -264.9754943847656, | |
| "loss": 0.6886, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.00148511934094131, | |
| "rewards/margins": 0.005857478827238083, | |
| "rewards/rejected": -0.007342599332332611, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.999733114418725e-07, | |
| "logits/chosen": -0.6464229226112366, | |
| "logits/rejected": -0.6577852964401245, | |
| "logps/chosen": -318.97540283203125, | |
| "logps/rejected": -295.6136169433594, | |
| "loss": 0.6858, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.002440870273858309, | |
| "rewards/margins": 0.01247528288513422, | |
| "rewards/rejected": -0.010034412145614624, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.990398100856366e-07, | |
| "logits/chosen": -0.7159032821655273, | |
| "logits/rejected": -0.6997270584106445, | |
| "logps/chosen": -257.1394958496094, | |
| "logps/rejected": -223.77249145507812, | |
| "loss": 0.6834, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": 0.008302798494696617, | |
| "rewards/margins": 0.01499713771045208, | |
| "rewards/rejected": -0.006694340147078037, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 4.967775735898179e-07, | |
| "logits/chosen": -0.6443796157836914, | |
| "logits/rejected": -0.6453703045845032, | |
| "logps/chosen": -287.23919677734375, | |
| "logps/rejected": -253.98779296875, | |
| "loss": 0.6809, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 0.012265635654330254, | |
| "rewards/margins": 0.02614836022257805, | |
| "rewards/rejected": -0.013882724568247795, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 4.931986719649298e-07, | |
| "logits/chosen": -0.6774781346321106, | |
| "logits/rejected": -0.6471685171127319, | |
| "logps/chosen": -281.8182067871094, | |
| "logps/rejected": -281.8425598144531, | |
| "loss": 0.6743, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.0038391489069908857, | |
| "rewards/margins": 0.02816380001604557, | |
| "rewards/rejected": -0.032002948224544525, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.883222001996351e-07, | |
| "logits/chosen": -0.7065152525901794, | |
| "logits/rejected": -0.692989706993103, | |
| "logps/chosen": -276.53753662109375, | |
| "logps/rejected": -264.0567932128906, | |
| "loss": 0.6657, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.04013616591691971, | |
| "rewards/margins": 0.060574889183044434, | |
| "rewards/rejected": -0.10071107000112534, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.821741763807186e-07, | |
| "logits/chosen": -0.6360586881637573, | |
| "logits/rejected": -0.6389707326889038, | |
| "logps/chosen": -301.81329345703125, | |
| "logps/rejected": -287.9957580566406, | |
| "loss": 0.6548, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.07365767657756805, | |
| "rewards/margins": 0.08726193010807037, | |
| "rewards/rejected": -0.16091960668563843, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "eval_logits/chosen": -0.7018380165100098, | |
| "eval_logits/rejected": -0.6576583981513977, | |
| "eval_logps/chosen": -290.2217712402344, | |
| "eval_logps/rejected": -294.29364013671875, | |
| "eval_loss": 0.6548193693161011, | |
| "eval_rewards/accuracies": 0.66796875, | |
| "eval_rewards/chosen": -0.1060016080737114, | |
| "eval_rewards/margins": 0.10316330194473267, | |
| "eval_rewards/rejected": -0.20916491746902466, | |
| "eval_runtime": 117.0477, | |
| "eval_samples_per_second": 17.087, | |
| "eval_steps_per_second": 0.273, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.747874028753375e-07, | |
| "logits/chosen": -0.6454014778137207, | |
| "logits/rejected": -0.5620900988578796, | |
| "logps/chosen": -347.7331237792969, | |
| "logps/rejected": -300.71112060546875, | |
| "loss": 0.651, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.11785294860601425, | |
| "rewards/margins": 0.12140049040317535, | |
| "rewards/rejected": -0.23925340175628662, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.662012913161997e-07, | |
| "logits/chosen": -0.70166015625, | |
| "logits/rejected": -0.727249801158905, | |
| "logps/chosen": -303.4729919433594, | |
| "logps/rejected": -263.29583740234375, | |
| "loss": 0.6384, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.18253406882286072, | |
| "rewards/margins": 0.14989741146564484, | |
| "rewards/rejected": -0.33243149518966675, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.5646165232345103e-07, | |
| "logits/chosen": -0.6562562584877014, | |
| "logits/rejected": -0.5979722142219543, | |
| "logps/chosen": -306.23516845703125, | |
| "logps/rejected": -346.08197021484375, | |
| "loss": 0.6379, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.28600436449050903, | |
| "rewards/margins": 0.16408249735832214, | |
| "rewards/rejected": -0.45008689165115356, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.456204510851956e-07, | |
| "logits/chosen": -0.5730563402175903, | |
| "logits/rejected": -0.5727633237838745, | |
| "logps/chosen": -335.1769714355469, | |
| "logps/rejected": -309.3208312988281, | |
| "loss": 0.6268, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -0.16110344231128693, | |
| "rewards/margins": 0.21982815861701965, | |
| "rewards/rejected": -0.38093167543411255, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 4.337355301007335e-07, | |
| "logits/chosen": -0.6809996366500854, | |
| "logits/rejected": -0.6591669321060181, | |
| "logps/chosen": -328.6540832519531, | |
| "logps/rejected": -292.2736511230469, | |
| "loss": 0.6047, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.33402004837989807, | |
| "rewards/margins": 0.2206876277923584, | |
| "rewards/rejected": -0.5547076463699341, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 4.2087030056579986e-07, | |
| "logits/chosen": -0.6660237312316895, | |
| "logits/rejected": -0.6344098448753357, | |
| "logps/chosen": -324.8284912109375, | |
| "logps/rejected": -335.0122985839844, | |
| "loss": 0.6017, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.30958321690559387, | |
| "rewards/margins": 0.312285840511322, | |
| "rewards/rejected": -0.6218689680099487, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 4.070934040463998e-07, | |
| "logits/chosen": -0.6780328154563904, | |
| "logits/rejected": -0.6576133370399475, | |
| "logps/chosen": -300.7780456542969, | |
| "logps/rejected": -299.3976745605469, | |
| "loss": 0.6237, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.4301220774650574, | |
| "rewards/margins": 0.2145705670118332, | |
| "rewards/rejected": -0.6446926593780518, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.9247834624635404e-07, | |
| "logits/chosen": -0.6824392080307007, | |
| "logits/rejected": -0.6258381009101868, | |
| "logps/chosen": -354.97503662109375, | |
| "logps/rejected": -370.0382385253906, | |
| "loss": 0.599, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.4686453938484192, | |
| "rewards/margins": 0.343492329120636, | |
| "rewards/rejected": -0.8121377825737, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3.7710310482256523e-07, | |
| "logits/chosen": -0.6199353933334351, | |
| "logits/rejected": -0.6015241146087646, | |
| "logps/chosen": -320.41021728515625, | |
| "logps/rejected": -331.60284423828125, | |
| "loss": 0.5902, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.27603933215141296, | |
| "rewards/margins": 0.31284210085868835, | |
| "rewards/rejected": -0.5888813734054565, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.610497133404795e-07, | |
| "logits/chosen": -0.7054450511932373, | |
| "logits/rejected": -0.6953166723251343, | |
| "logps/chosen": -338.31829833984375, | |
| "logps/rejected": -323.37115478515625, | |
| "loss": 0.5881, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.5281119346618652, | |
| "rewards/margins": 0.2860097885131836, | |
| "rewards/rejected": -0.8141217231750488, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "eval_logits/chosen": -0.8023231625556946, | |
| "eval_logits/rejected": -0.7415486574172974, | |
| "eval_logps/chosen": -310.1252136230469, | |
| "eval_logps/rejected": -339.1258850097656, | |
| "eval_loss": 0.5968395471572876, | |
| "eval_rewards/accuracies": 0.6875, | |
| "eval_rewards/chosen": -0.3050362467765808, | |
| "eval_rewards/margins": 0.3524515628814697, | |
| "eval_rewards/rejected": -0.657487690448761, | |
| "eval_runtime": 121.2666, | |
| "eval_samples_per_second": 16.493, | |
| "eval_steps_per_second": 0.264, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 3.4440382358952115e-07, | |
| "logits/chosen": -0.7312219738960266, | |
| "logits/rejected": -0.730713427066803, | |
| "logps/chosen": -314.7323303222656, | |
| "logps/rejected": -301.01312255859375, | |
| "loss": 0.5938, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.23817189037799835, | |
| "rewards/margins": 0.30586400628089905, | |
| "rewards/rejected": -0.5440359115600586, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.272542485937368e-07, | |
| "logits/chosen": -0.7227334380149841, | |
| "logits/rejected": -0.6831812858581543, | |
| "logps/chosen": -315.691162109375, | |
| "logps/rejected": -317.9109802246094, | |
| "loss": 0.5961, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.27428802847862244, | |
| "rewards/margins": 0.35738009214401245, | |
| "rewards/rejected": -0.6316681504249573, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 3.096924887558854e-07, | |
| "logits/chosen": -0.6957461833953857, | |
| "logits/rejected": -0.6545354723930359, | |
| "logps/chosen": -313.1252746582031, | |
| "logps/rejected": -355.5252685546875, | |
| "loss": 0.5924, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.4048747420310974, | |
| "rewards/margins": 0.39421346783638, | |
| "rewards/rejected": -0.7990882992744446, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.9181224366319943e-07, | |
| "logits/chosen": -0.8229473829269409, | |
| "logits/rejected": -0.8242266774177551, | |
| "logps/chosen": -316.42327880859375, | |
| "logps/rejected": -329.6378173828125, | |
| "loss": 0.5973, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.5616710186004639, | |
| "rewards/margins": 0.3975863456726074, | |
| "rewards/rejected": -0.9592572450637817, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.7370891215954565e-07, | |
| "logits/chosen": -0.7482638955116272, | |
| "logits/rejected": -0.7271997332572937, | |
| "logps/chosen": -325.5633850097656, | |
| "logps/rejected": -347.3030700683594, | |
| "loss": 0.5769, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.5051977038383484, | |
| "rewards/margins": 0.35222890973091125, | |
| "rewards/rejected": -0.857426643371582, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.55479083351317e-07, | |
| "logits/chosen": -0.8334507942199707, | |
| "logits/rejected": -0.8130620718002319, | |
| "logps/chosen": -333.03924560546875, | |
| "logps/rejected": -325.97979736328125, | |
| "loss": 0.5666, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.482491672039032, | |
| "rewards/margins": 0.45697951316833496, | |
| "rewards/rejected": -0.9394710659980774, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 2.3722002126275822e-07, | |
| "logits/chosen": -0.8548744320869446, | |
| "logits/rejected": -0.839653491973877, | |
| "logps/chosen": -324.4539794921875, | |
| "logps/rejected": -357.67169189453125, | |
| "loss": 0.5678, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.45763128995895386, | |
| "rewards/margins": 0.48610082268714905, | |
| "rewards/rejected": -0.9437320828437805, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2.19029145890313e-07, | |
| "logits/chosen": -0.814782440662384, | |
| "logits/rejected": -0.7766333222389221, | |
| "logps/chosen": -330.60662841796875, | |
| "logps/rejected": -326.7635192871094, | |
| "loss": 0.5829, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.46811041235923767, | |
| "rewards/margins": 0.3988150358200073, | |
| "rewards/rejected": -0.8669255375862122, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 2.0100351342479216e-07, | |
| "logits/chosen": -0.714547336101532, | |
| "logits/rejected": -0.656903862953186, | |
| "logps/chosen": -347.14349365234375, | |
| "logps/rejected": -363.637451171875, | |
| "loss": 0.5845, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.4874221384525299, | |
| "rewards/margins": 0.36368709802627563, | |
| "rewards/rejected": -0.8511092066764832, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 1.8323929841460178e-07, | |
| "logits/chosen": -0.7814041376113892, | |
| "logits/rejected": -0.7600786089897156, | |
| "logps/chosen": -350.5426025390625, | |
| "logps/rejected": -373.58282470703125, | |
| "loss": 0.5753, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.4952825903892517, | |
| "rewards/margins": 0.3200310170650482, | |
| "rewards/rejected": -0.8153136372566223, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "eval_logits/chosen": -0.8262501955032349, | |
| "eval_logits/rejected": -0.7578529119491577, | |
| "eval_logps/chosen": -328.7913513183594, | |
| "eval_logps/rejected": -371.0563659667969, | |
| "eval_loss": 0.5733634233474731, | |
| "eval_rewards/accuracies": 0.72265625, | |
| "eval_rewards/chosen": -0.4916972815990448, | |
| "eval_rewards/margins": 0.48509520292282104, | |
| "eval_rewards/rejected": -0.9767925143241882, | |
| "eval_runtime": 130.1337, | |
| "eval_samples_per_second": 15.369, | |
| "eval_steps_per_second": 0.246, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.6583128063291573e-07, | |
| "logits/chosen": -0.8206486701965332, | |
| "logits/rejected": -0.8467684984207153, | |
| "logps/chosen": -335.5056457519531, | |
| "logps/rejected": -316.3816833496094, | |
| "loss": 0.5633, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.5007157325744629, | |
| "rewards/margins": 0.4249514937400818, | |
| "rewards/rejected": -0.9256671667098999, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.488723393865766e-07, | |
| "logits/chosen": -0.9328675270080566, | |
| "logits/rejected": -0.8487465977668762, | |
| "logps/chosen": -306.1993103027344, | |
| "logps/rejected": -342.8832092285156, | |
| "loss": 0.5693, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.5110325813293457, | |
| "rewards/margins": 0.5352452993392944, | |
| "rewards/rejected": -1.0462777614593506, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.3245295796480788e-07, | |
| "logits/chosen": -0.6851174831390381, | |
| "logits/rejected": -0.6687122583389282, | |
| "logps/chosen": -352.14984130859375, | |
| "logps/rejected": -357.9172668457031, | |
| "loss": 0.57, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.570861279964447, | |
| "rewards/margins": 0.4501311779022217, | |
| "rewards/rejected": -1.0209925174713135, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.1666074087171627e-07, | |
| "logits/chosen": -0.7861064672470093, | |
| "logits/rejected": -0.7692282795906067, | |
| "logps/chosen": -313.29083251953125, | |
| "logps/rejected": -344.58062744140625, | |
| "loss": 0.5673, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.48139747977256775, | |
| "rewards/margins": 0.413718044757843, | |
| "rewards/rejected": -0.8951154947280884, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.0157994641835734e-07, | |
| "logits/chosen": -0.8523145914077759, | |
| "logits/rejected": -0.8958388566970825, | |
| "logps/chosen": -305.7944641113281, | |
| "logps/rejected": -335.16778564453125, | |
| "loss": 0.5533, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.4743496775627136, | |
| "rewards/margins": 0.49316510558128357, | |
| "rewards/rejected": -0.9675148129463196, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 8.729103716819111e-08, | |
| "logits/chosen": -0.7652881741523743, | |
| "logits/rejected": -0.8117391467094421, | |
| "logps/chosen": -384.6844177246094, | |
| "logps/rejected": -361.26385498046875, | |
| "loss": 0.5719, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.47557467222213745, | |
| "rewards/margins": 0.4706428647041321, | |
| "rewards/rejected": -0.9462175369262695, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 7.387025063449081e-08, | |
| "logits/chosen": -0.8511560559272766, | |
| "logits/rejected": -0.8533377647399902, | |
| "logps/chosen": -373.3354187011719, | |
| "logps/rejected": -363.8319396972656, | |
| "loss": 0.5718, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.5117169618606567, | |
| "rewards/margins": 0.5143001675605774, | |
| "rewards/rejected": -1.026017189025879, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 6.138919252022435e-08, | |
| "logits/chosen": -0.8471282124519348, | |
| "logits/rejected": -0.8318672180175781, | |
| "logps/chosen": -351.12689208984375, | |
| "logps/rejected": -389.46929931640625, | |
| "loss": 0.5572, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.6237812042236328, | |
| "rewards/margins": 0.5021282434463501, | |
| "rewards/rejected": -1.1259095668792725, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.991445467064689e-08, | |
| "logits/chosen": -0.8562029600143433, | |
| "logits/rejected": -0.8402010798454285, | |
| "logps/chosen": -342.40411376953125, | |
| "logps/rejected": -366.63250732421875, | |
| "loss": 0.5666, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.5931364297866821, | |
| "rewards/margins": 0.336722195148468, | |
| "rewards/rejected": -0.9298585653305054, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 3.9507259776993954e-08, | |
| "logits/chosen": -0.8323208093643188, | |
| "logits/rejected": -0.8248344659805298, | |
| "logps/chosen": -379.9523620605469, | |
| "logps/rejected": -394.05401611328125, | |
| "loss": 0.5602, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.5521629452705383, | |
| "rewards/margins": 0.45100274682044983, | |
| "rewards/rejected": -1.003165602684021, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "eval_logits/chosen": -0.92626953125, | |
| "eval_logits/rejected": -0.8578699231147766, | |
| "eval_logps/chosen": -329.260498046875, | |
| "eval_logps/rejected": -374.94903564453125, | |
| "eval_loss": 0.5689293742179871, | |
| "eval_rewards/accuracies": 0.734375, | |
| "eval_rewards/chosen": -0.49638885259628296, | |
| "eval_rewards/margins": 0.5193302035331726, | |
| "eval_rewards/rejected": -1.0157190561294556, | |
| "eval_runtime": 136.1974, | |
| "eval_samples_per_second": 14.685, | |
| "eval_steps_per_second": 0.235, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 3.022313472693447e-08, | |
| "logits/chosen": -0.9494141340255737, | |
| "logits/rejected": -0.9114382863044739, | |
| "logps/chosen": -325.3351135253906, | |
| "logps/rejected": -346.55963134765625, | |
| "loss": 0.5736, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.538487434387207, | |
| "rewards/margins": 0.4442841410636902, | |
| "rewards/rejected": -0.9827715754508972, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.2111614344599684e-08, | |
| "logits/chosen": -0.9237302541732788, | |
| "logits/rejected": -0.8802440762519836, | |
| "logps/chosen": -330.2814025878906, | |
| "logps/rejected": -381.2386169433594, | |
| "loss": 0.5607, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.5165634155273438, | |
| "rewards/margins": 0.437481164932251, | |
| "rewards/rejected": -0.9540446400642395, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.521597710086439e-08, | |
| "logits/chosen": -0.8156334757804871, | |
| "logits/rejected": -0.8552393913269043, | |
| "logps/chosen": -364.59478759765625, | |
| "logps/rejected": -368.82489013671875, | |
| "loss": 0.5777, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.5483056306838989, | |
| "rewards/margins": 0.5216010808944702, | |
| "rewards/rejected": -1.0699067115783691, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 9.57301420397924e-09, | |
| "logits/chosen": -0.9248915910720825, | |
| "logits/rejected": -0.8963875770568848, | |
| "logps/chosen": -329.3937683105469, | |
| "logps/rejected": -360.4046325683594, | |
| "loss": 0.5606, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.5845829248428345, | |
| "rewards/margins": 0.42287713289260864, | |
| "rewards/rejected": -1.007460117340088, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 5.212833302556258e-09, | |
| "logits/chosen": -1.0620511770248413, | |
| "logits/rejected": -1.0440218448638916, | |
| "logps/chosen": -308.4063720703125, | |
| "logps/rejected": -327.61328125, | |
| "loss": 0.5911, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.5545769333839417, | |
| "rewards/margins": 0.38095879554748535, | |
| "rewards/rejected": -0.935535728931427, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.158697848236607e-09, | |
| "logits/chosen": -0.9425494074821472, | |
| "logits/rejected": -0.9231334924697876, | |
| "logps/chosen": -308.18743896484375, | |
| "logps/rejected": -333.3728942871094, | |
| "loss": 0.5588, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.6228442192077637, | |
| "rewards/margins": 0.42768868803977966, | |
| "rewards/rejected": -1.0505329370498657, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 4.269029751107489e-10, | |
| "logits/chosen": -0.9223429560661316, | |
| "logits/rejected": -0.9098516702651978, | |
| "logps/chosen": -363.28765869140625, | |
| "logps/rejected": -387.8728942871094, | |
| "loss": 0.5479, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -0.480919748544693, | |
| "rewards/margins": 0.6743713617324829, | |
| "rewards/rejected": -1.155290961265564, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 478, | |
| "total_flos": 0.0, | |
| "train_loss": 0.6029723678173879, | |
| "train_runtime": 7448.3301, | |
| "train_samples_per_second": 8.208, | |
| "train_steps_per_second": 0.064 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 478, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 10000, | |
| "total_flos": 0.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |